Threatening or aversive stimuli normally evoke healthy fear, in which the brain?s defensive motivational systems drive protective behaviors. When a person or animal learns that a particular sensory stimulus predicts future harm, that stimulus begins to evoke fearful reactions as well. This is called learned fear. Sensory stimuli that are physically similar or conceptually related to the threat-predictive stimulus normally also evoke fear because of the reasonable belief that they might also predict future harm. This is called generalization of learned fear, and it is a normal part of healthy fear. However, many Americans suffer from disordered fear, in which the feeling of acute threat (fear) or potential threat (anxiety) generalizes to inappropriate stimuli and situations that may resemble or be associated with traumatic events but do not actually indicate an impending threat. Figuring out how to limit generalization so that a patient is only fearful in appropriate situations is a key practical challenge for the treatment of anxiety disorders. Most research on the biological basis of anxiety disorders and learned fear has focused on the mechanisms by which the brain learns the initial fear and extrapolates fearful behavior across situations. However, evidence is accumulating that learned fear also evokes profound changes in the brain?s sensory systems. This includes becoming hyper-sensitive and hyper-responsive to threat-predictive stimuli in ways that may underlie common post-traumatic symptoms like hyper-vigilance and attentional bias toward threat-predictive stimuli. In the previous project period we used a mouse model to observe how the neurobiology of the olfactory system is changed during appropriate fear and observed that fear learning makes the olfactory system selectively hyper-responsive to threat-predictive odor, generating a neural ?alarm signal? as early as the sensory input to the brain. In contrast, in this project period we will explore the more clinically- relevant situation of fear generalization, where initial trauma evokes fear of new stimuli that may not actually predict a threat. In preliminary experiments we employed an experimental paradigm in which mice undergo a traumatic experience associated with a particular odor but then generalize their fear across many novel odors. We will test whether this experience causes the olfactory system to become non-selectively hyper-responsive to many dissimilar odors, which might drive downstream responding as if the new odors are dangerous. We will also investigate the neural circuitry by which information about the traumatic event reaches the olfactory system and induces change in the response to many different odors. Finally, we will evaluate multiple candidate approaches to reverse the behavioral and sensory consequences of fear generalization, including comparing conventional exposure (a.k.a. extinction) therapy using the original trauma-associated odor with new therapy paradigms intended to ?refine? the generalized fear either through repeated exposure to novel stimuli or through follow-up fear training in which the trauma is explicitly paired with actually predictive stimuli but not with non- threatening stimuli.